 
|
Audio Asylum Thread Printer Get a view of an entire thread on one page |
For Sale Ads |
|
|
Audio Asylum Thread Printer Get a view of an entire thread on one page |
For Sale Ads |
68.126.21.44
The sound card in my Mac Mini is the Intel High Definition Audio 0x83847680 which does 192 kHz/32-bit quality for two channels, and 96 kHz/32-bit for up to eight channels. It comes standard, I've done no upgrading.
I can now confirm iTunes 8.0.2 does do 32 Bit music files, something no consumer physical format can do.
I compared the 96kHz and 192kHz sound samples from Acousence and here are my impressions. The 192kHz samples had more ambiance, notes have a longer decay, the bass is fuller and the strings are even smoother plus it sounds more live. I can "feel" the music on my skin almost as vividly as when I listen to live music.
I also compared these samples to my DSD recorded Exton SACD of Mahler's 6th with Ashkenazy and the Czech Philharmonic recorded in the House of Artist, Prague. The SACD and the 192kHz download were very close, the 192kHz download had more of the "live" feel on my skin but the percussion was more tingling exciting on the SACD, the string tone was actually smoother on the 192kHz. I would have to hear more examples but overall I think 192kHz music files may be sonically as good as or superior to SACD. I also have 192kHz PCM on DVD-Audios but I thought the music files easily beat those. I wonder if a DSD music file would also beat an SACD?
The 32 Bit 192kHz music file was a Jazz recording from Unipheye Music and sounded very real, there may be a further advantage to 32 Bit, I'd have to hear a few classical works before forming an opinion.
http://www.unipheyemusic.com/filetest.cfm
Some of these files are quite huge and I may be looking at external memory sooner than expected.
24 Bit 96kHz 4608 kbps WAV audio file
24 Bit 192kHz 9216 kbps WAV audio file
32 Bit 192kHz 12288 kbps WAV audio file
The 32 Bit 192kHz is almost 100 times larger than 128 kpbs MP3, that's a lot of memory. I will down the road need an external memory device that is whisper quiet as I live in a studio apartment and I understand many of them are noisy. Any ideas?
The performance of my little MAC mini was quite amazing considering I was using the analog outputs using a Stereo Mini to two RCA Monster Cable using one of the AUX inputs on my tubed pre-amp.
Give me high resolution or remain silent,
Teresa
Follow Ups:
1) Why do both iTunes and my MAC Mini's sound card, the Intel High Definition Audio 0x83847680 do 192kHz 32 bit PCM but the Audio Midi setup (core audio) only does 96kHz 24 bit? I would think MAC would at least try to match these devices up so they work in concert at the resolution that one actually downloads at. Or to put it another way, what good does it do to give iTunes the ability to do 32 bit 192kHz and use a sound card that does 32 bit 192kHz but limit the core audio to 24 Bit 96kHz. This does not make any sense to me whatsoever!
2) As you all know we learned that the Acousence Mahler 6th was recorded multitrack at 96kHz and then mixed to an analog mixing deck and then recorded at 192kHz. With the comparison files the 96kHz was a downsampled version of the 192kHz file. I downloaded both the 96kHz file (downsampled by Acousence from 192kHz) and the 192kHz which we learned my MAC mini downsampled to 96kHz. So why does my 192kHz files sound better than the 96kHz files when in the end they are both 96kHz?
3) I've listened many more times through speakers and headphones and the differences in greater bass, smoother sound, ambiance and longer decay of notes with the 192kHz files are very apparent. Why? I would think they would sound the same since the only difference is the 92kHz file was downsampled by Acousence and the 192kHz file was internally downsampled to 96kHz by the midi core audio.
I'm not complaining as all of these high resolution music files sound way better than 441kHz music files. It's just a lot of this computer stuff is weird.
Give me high resolution or remain silent,
Teresa
The 44khz sample from Acousence is terrible.
I know several of the best recording engineers in the world. None of them can here a difference between 96 and 192. In fact, the consensus seems to be that 48/24 is the practical limit of human hearing (where the improvement at 96 is extremely small).
The 192 file may be better, but it's unlikely due to the higher sample rate.
Teresa,
If there is no option for either 32 bit or 192 in Audio Midi Setup then you are basically listening to resampled music.
While it is true that iTunes can handle 32 bit material it will take that and set it to what every Audio Midi Setup is set to.
It will not output 32 bit since that is not a viable option in the setup.
Only 16,20 and 24 bit stereo at 44.1, 48 and 96.
Thanks
Gordon
J. Gordon Rankin
The highest setting on my Audio MIDI setup though the "Line In" (the only selection available) is 96kHz 24 Bit and that is where mine is set. Also my MIDI says it does not support Audio Output, I would guess because that is handled by the Intel High Definition Audio 0x83847680 sound card?
According to the Intel High Definition Audio 0x83847680 sound card in my MAC Mini it supports 192 kHz/32-bit quality for two channels, and 96 kHz/32-bit for up to eight channels.
What I downloaded off the internet in iTunes shows 32 Bit 192kHz 12,288 kbps WAV audio file and that is certainly large enough to indicate it is not downsampled to 96kHz.
And there is clearly an audible difference between 96kHs and 192kHz.
Give me high resolution or remain silent,
Teresa
Teresa,
You can set both input and output in Audio Midi Setup.
iTunes takes a snap shot of this when it loads and resamples the output of whatever track you have to this setting.
So you are probably downsampling everything at this point.
So YES you are limited to 24/96 on the output.
Thanks
Gordon
J. Gordon Rankin
Look at the drop downs in PROPERTIES FOR: in Audio Midi on the left
if you select built in input it says "output not supported" but select the ouput and you get options
if you select built in output it says "input not supported" but select the input and you get options
if you select an external device you get both input and output parameters
No 192kHz or 32 Bit although iTunes does and the Intel High Definition Audio 0x83847680 does 192 kHz/32-bit quality for two channels, and 96 kHz/32-bit for up to eight channels.
So am I getting all the resolution from my Intel High Definition sound card? Is the Audio Midi downconverting it to 192kHz to 96kHz and 32 Bit to 24 Bit?
Just played them again an the 192kHz music files sound different than the 96kHz music files.
Give me high resolution or remain silent,
Teresa
Higher resolution files are being downconverted. That's the difference you are hearing.
The only confirmed way to go above 24/96 is via an outboard dac via firewire (and optical, or so they say).
--eNjoY YouRseLf!.....
that the downsampled 192kHz music sounds better to you than native 96?
based on my experience in trying to put 24/96 music files in my iPod, they would not take the iPod refused them as the highest resolution the iPod will play is 44.1kHz WAV or AIFF. This is why I sold my iPod BTW.
I searched 192kHz, down converting and down sampling in Audio MIDI setup, iTunes and in my computer files, I don't believe I have any devices that have the ability to downsample. For example iTunes will not play 24/96kHz FLAC files I have to convert them first to 24/96 WAV files for them to play. I believe the same is true with 192kHz music files, since my computer will play them that can only mean I am getting 192kHz otherwise I would get an error message or dead silence.
The undeniable fact that the 192kHz music files clearly sound better as is an indication that I am getting the full 192kHz resolution. Could it be the MIDI set up device has not incorporated the 192kHz setting in their device and still plays the 192kHz files as is?
Give me high resolution or remain silent,
Teresa
iTunes will resample, either up or down, to the output settings in Audio Midi Setup. For example, if the output settings are set to 24/96 then a 16/44.1 file will be upsampled and a 24/192 file will be downsampled to 24/96. No errors, no silence, just sample rate conversion.This is a fact, not hearsay.
Ergo, you are NOT getting 192 resolution via the Mini's analog out and never will.
--eNjoY YouRseLf!.....
Edits: 02/12/09
So I guess I find 44.1kHz unacceptable even upsampled to 96kHz. But the 192kHz music files I downloaded sounded so much better than the 96kHz versions I am at a loss to explain it, especially the increased ambiance and the longer decay on notes. Could the increased ambiance be a byproduct of the higher jitter rate people say 192kHz has over 96kHz? Could the longer decay on notes be a byproduct of the downconversion? If so these byproducts are very musical?
So next question, since the sound card in my Mac Mini is the Intel High Definition Audio 0x83847680 which does 192 kHz/32-bit quality for two channels, and 96 kHz/32-bit for up to eight channels and iTunes does 32/192, if I got a different player would I then be able to play my 192kHz music files at the proper resolution?
Give me high resolution or remain silent,
Teresa
A different player will not let you play higher than 24/96. What you need is an outboard dac and a digital connection (preferably firewire, which I can confirm passes 192k material). The dac in the mini itself is of no consequence, as it's limited to 24/96 by Core Audio. Same is the case with the dac in the Airport Express and Apple TV: both are capable of high resolution multichannel decoding, but are limited to 16/48.
--eNjoY YouRseLf!.....
> based on my experience in trying to put 24/96 music files in my iPod,
> they would not take the iPod refused them as the highest resolution
> the iPod will play is 44.1kHz WAV or AIFF.
> This is why I sold my iPod BTW.
This is irrelevant to the question of whether your 192 KHz files were being downsampled.
> I don't believe I have any devices that have the ability to downsample.
How about the Core audio component of the Mac OSX?
> For example iTunes will not play 24/96kHz FLAC files I have to convert
> them first to 24/96 WAV files for them to play.
Converting Flac files at 24/96 to WAV files at 24/96 does not require downsampling.
> I believe the same is true with 192kHz music files, since my computer
> will play them that can only mean I am getting 192kHz otherwise I would
> get an error message or dead silence.
Reasoning without adequate knowledge is useless. Other people have tried to tell you something about Mac audio.
Bill
It is hard to tell with computer audio what is going on. This is a basic characteristic of computer software. A huge rats nest of connections can be programmed up and it will look exactly the same as a neatly organized structure to any user who doesn't have the tools and expertise to "look behind the curtain." With hardware things are a little more obvious, because you will likely see the rat's nest of wires or the neat wiring harness without needing any special tools (other than possibly a screw driver). (This is becoming less true as more and more components get shrunk into portions of a chip and the wires require microscopes and milling machines to be seen.)
If you want to know what is going on (For example, does a playback chain really run at 192 kHz?) you can do conduct a black box test. For example, you can play a test file with a sweep tone that goes from 20 to 96 K Hz and connect a scope to the DAC output and see what happens. If you don't have a scope, you can connect another computer with a 192 kHz sound card and use that computer with appropriate software as a scope
Don't trust anyone who claims that a system does or doesn't run at a certain rate, unless you know that (a) he has done these kind of tests and (b) your system is configured the same way his system. Good luck! Caveat Emptor!
In the case of the Mahler 6th symphony, I know that my DAC doesn't run above 96 kHz. When I try to play the 192 kHz files with Sound Forge I get an error message. However, I can play them with cPlay, because cPlay automatically downsamples the files to 96 kHz. As it happens, the 192 kHz files sound a little better than the 96 kHz files. Perhaps this is just a $3 placebo. Perhaps not. Consider that the 96 files were made by down converting from 192 in the production process. So when I did a comparison, what I was really comparing was different filters used in downsampling. It is well known that different digital filters can sound different.
Tony Lauck
"Diversity is the law of nature; no two entities in this universe are uniform." - P.R. Sarkar
d
Is that similar to Sudoku, with more colors?
Yes, it is a team affair with different colored teams (so now you can play defense making it a real sport). SOmetimes I forget what forum I am on :)
I just remembered. The captain of the team is Placebo Domingo. Fmak is the quality control manager.
I'm the cheerleader/announcer.
"I will down the road need an external memory device that is whisper quiet.... "
How about totally quiet?
You can try Solid-State Disk with USB that will plug right into your Mac USB port. Write speeds for the lower cost (MLC) SSD will be slow... but who cares? You're not going to spend most of your time writing to SSD.... you'll be reading your music files off the SSD when you play it back.
Of course, this does not address the spinning disk inside the Mac but you can replace that one with SSD too. (But the internal disk is pretty quiet).
Visit newegg.com, search on SSD, then list ALL SSD devices
The problem is that SSDs are not priced like regular hard drives (they cost much more per GB). My advice is to wait for rebates to show up on either the drive or size that you want and you'll save some $ (you could also go the ebay route, but some people don't like to do that). I purchased a SSD from newegg.com a while back and received a rebate for $50. That's not bad on a 64GB drive.
There's also slickdeals.net and buy.com, but some of the deals might only last a day or so (especially at slick). You have to be quick or you'll miss out.
I'm waiting a few more months for prices to come down even more, perhaps later this summer. My minimum requirement that I set for myself is 250GB and at that capacity, SSD is still WAY too pricey for my liking. If prices don't come down significantly by mid-summer, I may go 128GB instead.
Sounds like exactly what I need. My Mac Mini's disc is still quiet (knock on wood) though I have read about some have developing noise over time.
Give me high resolution or remain silent,
Teresa
Good advice, Abe.
24/96, 24/192, 32/192, 32/384. How about 64/768?
Or instead, how about recording engineers try some new methods to try and make better sounding recordings without abusing compression and every other effect known to man? And perhaps not being part of the silly "volume wars" game with a gazillion clipped samples? Then we'd see that recording, mixing and mastering skill outweighs sample rate as a function of sound quality by about 10 to 1. And that if recordings were all done WELL to begin with, we could have stopped happily at 24/96.
Remember too, that jitter only becomes more difficult to deal with as you increase the sample rate...
Sorry, just had to offer my usual cynical contratian wet-blanket view.
It's Monday after all...
Sorry. But I'll take awesome recordings at 16/44.1 over mediocrity at 24/192 or crap at 32/384 ANY. DAY. OF. THE. WEEK! Quality recordings in high res is a no brainer. But I'll pass on crap in high res. MP3 is fine.
Cheers,
Presto
"Or instead, how about recording engineers try some new methods to try and make better sounding recordings without abusing compression and every other effect known to man? And perhaps not being part of the silly "volume wars" game with a gazillion clipped samples? Then we'd see that recording, mixing and mastering skill outweighs sample rate as a function of sound quality by about 10 to 1".
This is all too common with popular music and I think a lot of it has to do with the artists and engineers thinking everyone is going to be playing the MP3 version.
The samples in these tests were Classical and Jazz and all superbly engineered and recorded. The best recordings sound even better as you raise the sampling rates. With the majority of poorly made current Pop especially those involved in the loudness wars, MP3 might just be the highest resolution for them. It seems the more you raise the resolution of poorly made recordings the more you reveal it's flaws.
Give me high resolution or remain silent,
Teresa
Teresa:
I think 24/192 is warranted with recordings like the ones you sampled.
I hope that the better rendering ability of the higher resolution formats does indeed incent recording engineers to put more effort into their methods.
I was just suggesting that we don't assume this will be the case if the "numbers are there".
I think I just need some tunes....!!!
Cheers,
Presto
You're presuming that all hi-rez recordings are crap, and will remain so in the future. The recording engineers that plan to utilize 24/176, 24/192, and certainly 24/384 are in-their-bones dedicated to the best sound quality they can muster. At least that's what I've heard and read up to this point.
There are at least two different issues, which have been discussed ad nauseum for years....great recordings with less-than-inspired performances (I'm being kind), and great music that was recorded poorly. You'll agree that there are far too many of those instances. And you would also agree that there is an overwhelming abundance of those at 16/44.
A contra for your contra.
nt
I'll drink to that.
Teresa,
Lately you have become a Mac maven. First a Macbook, then iPod, now a Mini. What is next, an iPhone?
George
I don't like cell phones and don't even answer my home phone, I let the answering machine do that.
Even though uncompressed lossless WAV files have more resolution than the higher bit rate lossy MP3 files they were too cold sounding. MP3s have an artificial warmth maybe due to the rolled off highs? At any rate in the end no 44.1kHz music file either at home or on the go was acceptable. So when I leave the house now it is without music.
Give me high resolution or remain silent,
Teresa
Nice review!
The only external storage that I have found that is quiet enough in my small apartment are notebook drives in external enclosures:
The size of notebook drives has increased tremendously up to 500GB now (e.g. Samsung). I use several external Macally PHR-S250CC enclosures .
Luckily the Man Mini is also very quiet...
Die gefährlichste Weltanschauung ist die Weltanschauung derjenigen, die die Welt nicht angeschaut haben.
Alexander von Humboldt (1769-1859)
Give me high resolution or remain silent,
Teresa
"I also have 192kHz PCM on DVD-Audios but I thought the music files easily beat those."I have to believe, and I have no proof of this yet, that removing the lazer mechanism, and the mecahnical parts associated with reading an optical disc somehow does result in benefit to sound reproduction.
I would imagine the same holds true for DSD played back from a hard drive vs an SACD player.
I am sure someone on these forums might have that experience and could shed light on it.
navman
Edits: 02/09/09
And by fate I ordered 2 of the Acousance 24/192 releases last week. Looking forward to them.....
Regards,
Joel
I am disappointed that Acousence records at 192k, then edits in the analog domain and revert back to digital. They need to invest in a digital workstation.
I have ordered the Mahler 192k and will playback on the ESS Sabre.
You are right, 192k sounds better than 96k. DSD digital out from a sacd player into a dsd dac is even better; it does not have the hf sonic signature of pcm.
"You are right, 192k sounds better than 96k. DSD digital out from a sacd player into a dsd dac is even better; it does not have the hf sonic signature of pcm."
Are you sure the sonic signature of 192 kHz PCM is due to the format and not the equipment used in making the recordings or that you are using for playback?
"I am disappointed that Acousence records at 192k, then edits in the analog domain and revert back to digital. They need to invest in a digital workstation."
I read that Acousense web page, too. It makes no sense to me. I can see doing it if they are playing fancy games in real-time with the mix or performing a lot of EQ or other processing, but I don't think that either would fit well with their philosophy. DAW software that will run on an under $1000 PC would cost less than a single lunch for a symphony orchestra. I wonder just what processing they are doing in the extra pass.
Tony Lauck
"Diversity is the law of nature; no two entities in this universe are uniform." - P.R. Sarkar
"I am disappointed that Acousence records at 192k, then edits in the analog domain and revert back to digital. They need to invest in a digital workstation."
Hi,
I´ve just found this thread and it seems that there is some confusion about the workflow during the recording process at ACOUSENCE. Perhaps it is described not clear enough on our webpage. So please let me explain some things.
We record multitrack @96k with a DAW (and two HD recorders as backup), we mix the multitrack recording down to 2-track with an analog mixing desk and then we record the stereo mix @192kHz. Sure, we could do the downmix inside the DAW, but the analog mixdown is a very, very important thing to get a really convincing musical performance. It is a big misbelief if someone means that doing all in the digital domain is always the best. It is the cheapest, that´s right.
We need 24 or more highclass DA converters, a very expensive mixing desk and two highclass AD converters. But in my mind, a digital mix (especially inside the DAW) would be some of the worst things we could do with our recordings.
Thanks, Teresa for your interest in our recordings and your detailed description. If I read such things like "... it sounds more live", I see that it is good that we invest the additional expenses. This workflow is one brick to get this "live performance".
Best regards
Ralf Koschnicke
ACOUSENCE records
Then I am afraid that you cannot imply selling a 192k recording. The price should also not be set at a premium for this sample rate/
Musically and soundwise I shall reserve judgement until I receive the discs.
I have no bad conscience with selling 192k recordings. That´s the reason why I make no secret about the workflow. The important point is, that the analog stereo mix is recorded at 192kHz. It would be impossible to archive the same quality @96k and so you get a real benifit of our 192k version. Eyeryone who owns both versions can hear it.
The other thing is that each recording track for only one microphone gets no real benefit from using more than 96k. We checked this very precise.
At this question the analog mixing is the key again, but to explain the WHY in detail would be very complex.
All in all you get the main benefit of high sample rates when switching from 44,1/48k to 96k. The 192k adds a additional slight benefit. Slight but not unimportant, I describe it like: Now I get the last little bit for sounding really analog. But the benefit does very much depend on your system. Handling 192k really good is no easy thing in the hifi and/or computer world, so my experience.
So you also find this in our pricing. 192k recordings compared with 96k are only a little bit more expensive, not as much as the additional DVD costs or the data handling of the doubled file size.
regards
Ralf
1) You said you recorded the music at 96kHz in multichannel
2) Mixed it to analog Stereo
3) Recorded the analog mix to 192kHz
4) How was the final 96kHz mix made? Was it directly from the analog mix or a mix down from the 192kHz PCM?
I compared the 96kHz and 192kHz music files, and clearly heard a sonic advantage of the 192kHz music files over the 96kHz music files. But I am confused now as to why, since the original master is 96kHz, could it be the added warmth from mixing in analog requires 192kHz to capture all of it?
I have no problem with you mixing to analog, I am just curious why a 96kHz original mixed to analog should sound best at 192kHz? Also why 192kHz isn't needed for the original when it sounds best on the final mix?
Thanks in advance,
Give me high resolution or remain silent,
Teresa
I shall listen
The reason why I made the remark is that I prefer to have hirez files of known provenance (as in food!). This allows me to judge in clearer terams how my equipment performs at different sampling rates and not reproduce the sonic signature of other peoples' equipment (which isn't always good).
Your upsampling may make the 192k files sound better but I would prefer to have the native 96k ones, and do my own software and/or hardware upsampling.
I don't mind people charging more for clarity which at least you have. It is not the case for some others..
We are not talking about a simple upsampling process. I know that there are many hirez files on the market that are only upsampled; you can see this in a FFT spectrum. So I understand your wishes, but please let me explain it in an other way:
The end product exists at this moment when the signal exists on the stereo output of the mixing desk. All steps before are parts of the whole. Each component is chosen to get the best possible sound and music. The recording format is only one thing. The final sound of the recording is a result of all this technical components and of the creativity and capabilities of the recording producer.
Music recording is not as easy as mathematics or bookkeeping. The technical dates are no guarantee for the result. A absolute "no go" are samplerates below 96k because then we really lose details. The limiting factor are the microphones. They deliver signals that you cannot archive @48k, but they deliver no signal you could only archive with 192k. So 96k is the right samplerate for mono microphone tracks. Also the ringing effects and energy dispersion are large @44,1/48, but low @96k and only a bit lower @192k. At this point the quality of the converter is much more important than the question 96k or 192k. Above 96k there are other effects more important. So it is very easy to get a worse sounding 192k recording because of bigger problems with jitter and DAW timing.
But now during analog mixing you get a much more complex mix of the signals and this complex mix results in a signal which contains much more details than each of the recording tracks. Now 96k is not enough to archive this complex mix.
Perhaps I have underestimated the problems in communication about this point. So it is good to talk about. It would be much easier to tell the people that we use 192k at all steps, but why should we do this when the result is not better or even worse. We own 24 AD-converters @192k, but the first tests with 192k multitrack have been not really convincing.
So if you listen to the 96k version of our recordings this is not more close to the original. 192k ist the original, all other formats are downsampled from this. Not the reverse way.
... I just remembered on one example:
You can compare this workflow with the old analog studio workflow. Multitrack recording on 24 track 2 inch, Mastering on 2 track 1/2 inch. No one built a 24 track machine with 1/4 inch for each track but many people used 1/2" master machines for highest quality at the stereo mixdown.
Well if the original has been sampled at 96 khz then there is no information to be gained by sampling at any higher rate. Even if you mixed all 24 bit tracks into a single track you would pick up at most 5 extra bits of resolution. So if you used a digital mixing deck that had at least 32 bits (fixed point) or 64 bits (floating point) resolution, you would be able to capture and mix everything that your first level of recording has captured.
Now when you dither this 29 bit resolution down to 24 bit, still at 96 kHz you will lose a certain amount of resolution, although if you use dither and noise shaping the loss will be slight. If you upsample to 192 and then use dither and noise shaping to reduce down to 24 bits you will be able to use more noise shaping and hence there will be more resolution in the final product. But the difference is going to be extremely slight.
When you convert out to the analog domain and then back into digital you add in the dispersion caused by two separate filters, one in the DAC and one in the ADC. This can't do anything to improve the sound. This has to be a much greater degradation than any minute gain in low level resolution below -145 dbFs. It is possible that there is euphonic distortion in the extra converters or in your analog mixing deck. (This would be true if you were using plug ins that had no equivalent digital version, but you didn't mention any.)
I can well believe that the output of your mixing deck sounds better than the output of a DAC on a digital audio workstation you tried. However, this is not the issue. The issue is whether the output of the consumer's DAC sounds better done one way or the other. To evaluate this you may need better monitoring equipment. It is also entirely possible that any DAW you tried had improper processing. It is not uncommon to have buggy DSP software.
If you still have all of the original 24 tracks of the Mahler 6th and accurate documentation of all of the mixing settings used, I would wager that it would be possible to produce a better sounding version of this recording, one that sounds even better than the existing 192 versions, but at 96 kHz.
Tony Lauck
"Diversity is the law of nature; no two entities in this universe are uniform." - P.R. Sarkar
We should be careful that we don´t discuss on a theoretical basis without any value for real life. The whole recording process consists of a sophisticated combination of technical tools and methodical considerations. The target is an artistic result. I tried many different ways and I have good reasons for exactly going this way. It is absolutely impossible to transfer files and parameters on an other system and getting a comparable result; perhaps better, perhaps worse, but in any case not the same artistic result. This is not the way music production works.
I think we can conclude that we can´t easily answer the question about the sense behind 96k or 192k. But outgoing from my final result at the ouput of the mixing desk it is only possible to archive exactly this final result with a ADDA chain @192k. 96k is quite good, but I hear a difference if I compare the direct mixing desk output with the ADDA chain. That´s fact. And: The result of downsampling to 96k from this 192k master is more close to the original than a 96k AD converter signal. I checked such things very accurately before I decided to work in this way. So you may be certain that with our offered 192k version you get the best possible version of our recordings. The master @96k is quite good, too, but it is not really the final result I got out of the mixing desk; even if the original recording was done @96k.
Please accept that a combination of many transmission channels results in a signal with a much more higher density of information. Please remember to my example with 2" tape for multitrack recording and 1/2" tape for mastering. Before establishing the CD format the music industry never worked with identical resolutions at each production step. That´s a very curious "achievement" of the digital era. But the practical experience shows that this old principle is still right. Today there is only one difference: It is not necessary to reduce the master quality for a consumer media, we are able to transfer the real master quality to the end user. That´s fine, isn´t it?
"We should be careful that we don´t discuss on a theoretical basis without any value for real life. The whole recording process consists of a sophisticated combination of technical tools and methodical considerations. The target is an artistic result."
If the final artistic result were less than excellent, then your argument would be purely theoretical, also. But the result was excellent, and that's what counts. :-)
Thanks for participating in the discussion.
Tony Lauck
"Diversity is the law of nature; no two entities in this universe are uniform." - P.R. Sarkar
Thank you for this discussion, too. And thank you for your interest in our recordings, fine that you like it.
Best wishes
Ralf Koschnicke
*** If you still have all of the original 24 tracks of the Mahler 6th and accurate documentation of all of the mixing settings used, I would wager that it would be possible to produce a better sounding version of this recording, one that sounds even better than the existing 192 versions, but at 96 kHz. ***
I'm not too sure. Ralph described a technique that was quite common in the 80s and early 90s - record digitally, then mix in analog, and reconvert the mix back to digital.
The main advantage of this technique is reducing the quantization errors caused truncation during the mixing process. The analog mixing desk also "dithers" the signal by adding noise.
Potentially we could achieve a "better" (or at least "cleaner") result by mixing in 64-bit fp (rather than 24-bit fixed or 32-bit fp which causes the quantisation noise), but there aren't too many of 64-bit DAWs around (I use SONAR, but I don't know of too many pros that use it).
Presumably the original intention of capturing at 96K was to avoid the typically poor(er) characteristics of 192kHz ADCs, but then the final result should also have been recaptured at 96kHz.
So in the end I kind of agree with you - given the source material is at 96kHz, the final master should also be at 96kHz (for the same reasons why the source was captured at 96kHz).
Given that we now understand the workflow, it would seem plausible that differences between 96kHz and 192kHz versions of this recording can also be explained through:
- quantization noise in the 96kHz version due to resampling from 192kHz
- differences in performance characteristics of the playback DAC at these two sample rates
Well in the unlikely event that Ralph gives me files and settings, I'll be glad to give it a try. :-) My DAW works at 64 bit floating point. But it's not set up to do serious mixing, so it would be tedious if there are any changes to the gain settings during the music.
Today I downloaded another of Ralph's recordings, the Wagner. Like the Mahler 6th it is an excellent recording.
Tony Lauck
"Diversity is the law of nature; no two entities in this universe are uniform." - P.R. Sarkar
Keep up the good work, and setting the "record" straight.
And some of what happens in the studios in general nowadays , because of digital itself, or poor training leads to aweful sounding CDs.
Would you say, that mixing in digital brings about problems? And if so, what problems are they?
navman
I'm beginning to think EVERYTHING about this hobby is an art. It's hard to even get the EXPERTS to agree on anything.
I believe it is the brickwall filter and the subsequent analog filter that causes much of the difference. Afterall, one is dealing with very high roll-of slopes and massive phase shifts introduced by these. There is sufficient measured and published historical data to support this. On my dCs 954, the type of digital filter has a major influence on the sound. This is also true of CD/SACD players that have various filters built in, such as the Sony SCD777ES. Also see Charles Hansen's latest white paper. There are earlier ones as well.
I hear clear differences introduced by diffferent Fcs in even single pole filters feeding the analog stage.
Don't tell Thomas Pf....he thinks we're imagining it.
"Don't tell Thomas Pf....he thinks we're imagining it."
In some philosophies, the Universe exists in the imagination of God.
Tony Lauck
"Diversity is the law of nature; no two entities in this universe are uniform." - P.R. Sarkar
Quite the opposite. I congratulate you and of course will be happy to do a DBT on your system :-)
Cheers
Thomas
Well, when I compared 320k MP3 to 16/44.1, I just couldn't tell a difference on rock, however, when I started to compare with my classical collection, it was easy to tell a difference for sure. I can't wait to compare 16/44.1 - 24/96 - 24/ 192.
Were you able to obtain statistically valid ABX results when you did the comparison?
I'm curious, because on hydrogenaudio, at bitrates above 300kbps, no one has been able to show they can distinguish between lossless and lossy.
I can distinguish between 192kbps and 256kbps, only just, and even I gave up between 320kbps and lossless. And I can normally hear differences between different dithering algorithms.
As for 44.1 vs 96 vs 192, I do hear a difference, mainly caused by reduced ringing effects at higher frequencies. I don't claim to hear any difference due to ultrasonics.
I think I personally would have a hard time distinguishing 320 kbps from lossless, because although the degradation has a different character relative to uncompressed, it's subtle enough to be difficult to recognize one from the other.
I can hear a difference between 44.1 kHz and 96- There is a slight improvement in HF "air", but on the other hand, it's not nearly as listenable. (If they can make 24/96 or SACD listenable, I'd buy such recordings.) I could *not* distinguish 24/96 from 24/192, aside from listening for longer than 10 minutes, 24/192 is even less listenable than 24/96.
I've brought up the test I've been running between a CD-to-wav file and a CD-to-320kbps MP3-to-wav file. Most people have preferred the MP3-converted file over the CD quality file, and whenever people guessed which was which, there have been more wrong guesses than correct ones.
....are more easily discernible with better playback software.
You were mentioning the Izotope, yet on that site, I could not decipher which one was to be used for playing back audio files..... Can you point me to it? Thank you!
> You were mentioning the Izotope, yet on that site, I could
> not decipher which one was to be used for playing back
> audio files..... Can you point me to it? Thank you!
Ah, must be Izotope Vinyl.
http://www.izotope.com/products/audio/vinyl/
;->
Seriously, though, Ozone 4 is a plugin for a DAW host
like WaveLab (and even Ozone MP is a plugin for something like Winamp
or WMP).
So maybe he's using the standalone version of Izotope RX
(the audio restoration package) as a player.
Izotope RX Advanced includes their "64-bit SRC",
and yes, it's expensive -- $1200.
Recommended and installed/configured by Vincent of VRS. If one thinks counter-intuitively, the purchase of this software and any less-expensive DAC/interface will reap great sonic rewards. The audio engine (as Vincent describes it) is far more important than many realize, especially since most of us have been conditioned to be gearheads (Ooooohh....a new DAC on the market!!). Purchasing a very expensive DAC/interface, but using JRiver or Samplitude or one of the Foobar versions reaps far less.
In comparison, Izotope RX makes JRiver, et al, sound like cardboard....I kid you not. It actually makes far more "sonic sense" to spend $1200 on this software than thousands on a DAC/interface. One can always upgrade the DAC down the road, but using an inferior playback engine with an expensive piece of gear is limiting the ability of the gear....no different in concept than using cheap cabling with an outstanding preamp or amp. Plain and simple.
Izotope RX is the only product that they make that has an audio engine. For me, it is by far the best audio engine on the Windows platform. But as you said, expensive, but well worth it.
What is an audio engine? This has no meaning and if guys stop talking like this and delineate clearly the differences, then we shall all be much clearer on what is going on.
I believe the term 'audio engine' refers to the playback portion of the software. I've heard this term a number of times over the last couple of years, and not just from Vincent. He can correct me if I'm wrong, but I think it's "developer-speak".
The reason why I queried this is the statement that it is the only software with an audio engine.Tthis can mean loads of things and if it is the player, how is it different from Audition, Samplitude , or whatever, and why does it sound better.
By using undefined terms like this, the post does not serve to inform and may misinform those who just read and then repeat it.
> The reason why I queried this is the statement that
> it is the only software with an audio engine.
The statement was "Izotope RX is the only product that
they make that has an audio engine." where "they" clearly
refers to Izotope.
The context of this entire thread makes it perfectly clear
what's being discussed. Krieger wanted to know which
Izotope product Kafton was using as a player. I looked
at the site and guessed it must be RX, because that's
the only thing that's capable of being used as a
stand-alone player. Sanders replied that yes, RX is
the only product they (Izotope) make that has an audio
engine (i.e., is capable of being used as a stand-alone
player).
Where's the beef?
nt
Alan,
Perhaps you meant “dialectric”?
I know what “dialectic” means and I am sure it does not apply here as there was no argument or even difference of opinion - just a request for information. Anyhow, I too was unaware of what an “audio engine” was. I had googled it and found a range of products promising “great sound”. Clearly, this was not what was being referred to in this discussion. Fmak asked the question that was certainly on my lips. Now I know and thanks to all. Clarification is a beautiful thing.
> Perhaps you meant “dialectric”?
You mean "dielectric"? ;->
Tony Lauck
"Diversity is the law of nature; no two entities in this universe are uniform." - P.R. Sarkar
....it's a dialogue. ;--0
Jim, I am not sure what I mean anymore but thanks for the clarification nonetheless. Apparently in your part of the world "Where's the beef?” amounts to a “clear conclusion”. In my part of the world it means that it is getting near dinner time.
> Apparently in your part of the world "Where's the beef?”
> amounts to a “clear conclusion”.
I just meant (semi-humorously), "What's the complaint?"
No....I meant dialectic. As I learned from my philosophy professor back in the day, argument and/or opposing forces are not always necessary to create a dialectic. + and + can = + . Jim beautifully stated certain facts that led to a clear conclusion, thus I called it a dialectic.
And yes, indeed, clarification is a beautiful thing. :--)
Playback engine is a term used by guys like Sonic Studio.
Edits: 02/10/09
I don't understand. You said 96k had more HF 'air' than 44.1k yet 44.1k was more listenable. I prefer an increase of 'air' between instruments but I am guessing you mean something different when you say, "HF air". Why would more 'air' be less desirable?
I don't claim to speak on behalf of Todd (although I have been accused of doing so in the past - not by Todd though). Take the following as ramblings from an idle mind.First of all, some people actually find PCM ringing at 44.1kHz euphonic. It may add a layer of complexity and thickness to the sound. People who object to ringing usually are objecting to "pre ringing" (since it anticipates the actual impulse) as opposed to "post ringing" (which may sound a little bit like hard reverb, which is generally regarded as euphonic)
Incidentally, if you want to train your ears to detect pre-ringing, try listening to a recording of percussion, like Taiko for example. On 44.1, you get this "sixth sense" feeling that you can anticipate exactly when then next drum beat will sound (it's because your ears are picking up the pre-echoes).
Secondly, on most DACs, THD and IMD is higher at 96 compared to 44.1/48 due to the design of the DAC. The higher distortion should be just audible if you have a really good system. At 192kHz, most DACs don't really perform all that well, based on my reading of their data sheets (and confirmed by actual measurements).
It's much harder to train your ears to detect higher IMD and THD, primarily because most systems don't have enough resolution due to various factors. It usually comes across as listener fatigue - after a while, repeated spectral patterns due to distortion artefacts can affect our perception of sound.
The trade-off (between higher distortion and less ringing) depends on your listening preference. Choose your poison.
Edits: 02/08/09
Try these tests to see if you can hear pre-ringing.
But note: if your DAC uses linear phase filtering then its pre-ringing may mask the differences between the files.
Tony Lauck
"Diversity is the law of nature; no two entities in this universe are uniform." - P.R. Sarkar
Yup, that sounds about right. You hear a difference but it might be due to higher distortion and noise.
Cheers
Thomas
Nah, I didn't do a valid DBT. Also, now that I think about it, I'm not sure it was 320k. It was years ago. Maybe I should do another test. One thing I am absolutely sure of is I can definitely hear the difference between 128k AAC, 192k MP3, and 320k MP3 IR broadcasts (I prefer 128k AAC over 192k MP3 but prefer 320k MP3 over 128k AAC). But, this really isn't a good comparison either because the feeds could be manipulated before they are broadcast. Anyway, I can't wait to have fun comparing 16/44.1 - 24/96 - 24/192 once I get a 192k capable DAC.
You can try this test, which incidentally will also train your ears to be able to distinguish lossy artefacts.
Take a music piece that you know really well, with lots of complex high frequencies (cymbals, steel brushes, bells etc.)
Convert them at a variety of bitrates (64, 96, 128, ... all the way to 320).
First listen to the lossless, then listen to the 64. You should be able to clearly hear the artefacts. Now progressively alternate between lossless and the next higher bitrate until you can't hear the artefacts.
For me, the cutover point tends to be around 192-256 kbps.
Post a Followup:
| FAQ |
Post a Message! |
Forgot Password? |
|
||||||||||||||
|
||||||||||||||
This post is made possible by the generous support of people like you and our sponsors:
Top of Page ]
[ Contact Us ]
[ Support/Wish List ]
[ Copyright © 2024, Audio Asylum, all rights reserved ]
